1. Efficient Market Theory

ISSN 1822-8011 (print) ISSN 1822-8038 (online) INTELEKTINĖ EKONOMIKA INTELLECTUAL ECONOMICS 2011, Vol. 5, No. 4(12), p. 602 612 January Effect at the Czech Capital Market Jan Luhan, Veronika Novotná, Vladěna Obrová Brno University of Technology, Faculty of Business and Management, Department of Informatics, Kolejní 2906/4, Brno, Czech Republic Phone No.: +420541143717, +420541143718, +420541143709 E-mail: luhan@fbm.vutbr.cz, novotna@fbm.vutbr.cz, obrova@fbm.vutbr.cz Abstract. In the course of the investment process it is necessary to examine a large number of stocks (stock groups) within wide categories of financial assets. The main objective is to find stocks that are not rated correctly at the given moment, appearing thus from the buyer s point of view as interesting to buy. To make such an analysis there is a wide scope of approaches. One of them is the efficient market theory. The main aim of this article is the application of methods of efficient market tests on the Prague stock exchange in the period of 2007-2010. In this article the authors aim at verifying the existence of the January effect at the Czech capital market, particularly in the conditions of the Prague Stock Exchange. The primary data used herein were obtained from official lists of prices of the Prague Stock Exchange. JEL classification: G14, C12. Keywords: Capital market, stock exchange, prediction, January effect, weekend effect, verification. Reikšminiai žodžiai: kapitalo rinka, vertybinių popierių birža, sausio efektas, savaitgalio efektas, patikrinimas. 1. Efficient Market Theory The first references about the efficient market can be found in articles by Bachlier from 1900. His work also explained the theory of random processes of commodity prices, being entirely original but in its time it was not given too much attention. Furthermore, it appeared in an empiric survey carried out by Cowles in 1937. However, Samuelson was the first to outline the concept of the modern of efficient market (Samuelson, 1973) A comprehensive theory of efficient markets was formulated for the first time by Eugene Fama in the 60s and 70s of the previous century, summarising the theory known until then and above all, the empiric results available. He then implanted the definition of efficient capital markets in the simplest form into the general framework

January Effect at the Czech Capital Market 603 of efficient market in economic terms, i.e. a market where prices fully reflect all the information available (Fama, 1991). 1. Fama s concept was followed by many renowned financial economists such as R. A. Brealey, S. C. Myers, J. C. Francis or R. A. Haugen, who defines the efficient market as a place where stock prices reflect all the information that is possible to know and that is significant ( Haugen, 1996). Fama himself later used a more general definition deeper set into economic terms where prices reflect the information available to the extent that the limit benefit produced by usage of the information is equal to the cost of acquisition thereof (Fama, 1991). Malkiel defines capital markets as efficient if buyers are not able to achieve excessive returns without accepting excessive risk (Malkiel, 2003) 2. Basic Assumptions The efficient market theory is currently the most significant stream in economics. The theory says that the stock market is efficient and price-creating information is contained therein nearly immediately (relevant information spreads in 30 seconds as maximum). This is the reason why business strategies fail because the price is always objective and adjusts to its internal value, and price movements are affected only by unexpected information, hence the price changes unexpectedly, too. Efficient market assumptions are: There is a huge number of rational buyers at the stock market continuously analysing and trading. Buyers have got sufficient information that is true, cheap and current. Buyers respond to new information accurately and quickly. Transaction costs are low. The market is near-money. None of the participants is in any monopolistic or exclusive position. Quality infrastructure and legal market regulation. According to the efficiency power we distinguish three basic forms: 1) Weak efficiency form the price contains historical information and therefore price predicting based on the previous price curve is not possible, which casts doubts on the technical analysis. 2) Semi-strong form the price contains historical and current public information, which casts doubts on the fundamental analysis, too. 3) Strong efficiency form the price contains any and all information, including non-public one. Many researches dealing with the efficient market theory have been carried out but only the weak and semi-strong efficiency forms were confirmed. It was confirmed that there is non-public information accessible to elect professionals only who thus achieve above-average revenues. This fact is also confirmed by Liu Yongxin in his study of 2009 (Liu, 2009).

604 Jan Luhan, Veronika Novotná, Vladěna Obrov The efficient market (EMH) has been applied to many other markets. Very little research has been done on electricity market efficiency testing. In their paper The Efficient Market Hypothesis and Electricity Market Efficiency Test the authors analyse the characteristics of an electricity market, comparing with other markets, and propose a testing approach for electricity market efficiency assessment (Zhe, 2005). The efficient markets (EMH) states that at any time, the price of a security fully captures all known information about that stock, so the price behaves like a random walk in time, except when there are changes in information. The authors of the Back propagation as a test of the efficient markets test whether a non-linear statistical method, error back propagation, can do better than chance in forecasting stock trends. The paper presents some research on the application of artificial neural networks to economic modeling (Tsibouris, 1992). Efficiency tests are divided in the same groups as the forms thereof. Therefore, Fama in his later work again introduced three groups, but with modified names, giving a better description of the testing procedure (Fama, 1991). The new test groups are revenue prediction tests (for the weak form), case studies (for the semi-strong form) and private information tests (for the strong form). In order to examine the semi-strong form of efficient capital markets we must first of all check the basic assumption whether the markets are at least weakly efficient. 3. Revenue Prediction Tests This group of tests includes a large number of testing procedures. They may be divided into two main groups, which, it needs to be noted, are very closely linked with each other. The first group contains tests of fair game and random walk, including randomness tests, run test, variance ratio test and examination of series correlations. The other contains anomalies, i.e. regularities in the development of stock returns difficult to be explained in economic terms. One of the anomalies is the January Effect when the returns of the first days of January and the whole month of January in total tend to be significantly higher, compared to the other months. Another of the anomalies is the Weekend Effect when Monday revenues tend to be significantly different, compared to the other days. Parallel to the Weekend Effect is the National Holiday Effect when the returns of the day following the day off (nontrading) tend to be different. 4. Market Anomalies Market anomalies are specific features in the stock behaviour that cannot be simply explained on the statistical basis, and sometimes they are difficult to explain on the economic basis as well. A frequent solution of such problems is an approach explaining

January Effect at the Czech Capital Market 605 anomalies on the basis of specific features in the behaviour of economic players. It is really hard to explain in economic terms why the stock should trade more on Wednesday than on any other day. Most of the studies accomplished so far primarily document the situation on the American market. Papers examining the efficiency level in other countries appear only sporadically and if they do, they usually refer to highly developed countries, such as Great Britain, Germany or Japan. Notwithstanding that fact, there is a study dedicated to the Czech capital market, testing the efficiency thereof, namely the one by J. Filáček, M. Kapička and M. Vošvrda of 1998. The authors came to the conclusion that the Czech capital market does not even reach the weak form of efficiency based on their testing of 1995-1997.(Filáček, 1998) In our previous article (Luhan, 2011) we drew up an analysis of one of the effects mentioned above the Weekend Effect. The Day-of-the-Week Effect was confirmed in the analysed stock, although not in a clear way. Unlike the world markets, where in most cases there was a significant negative return on Monday, on the Czech stock market we observed a significantly higher return on Monday and significantly lower return on Friday if we examine only average returns per day. However, if we study the anomaly in more detail and try to find the reason of it, we will first come to the conclusion that lower Friday returns can be caused by a psychological effect. This effect shows that in the Czech Republic the buyers working activity on the last days of the week is not as high as on the other days. The Day-of-the-Week Effect shows itself on different days in each of the years studied, in some cases it is not present at all. Using an investment strategy based on the observed anomalies, the buyer would not reach any excessive return in the next year. 5. January effect The January effect is one of the most observed and studied anomalies. It would be more appropriately called The effect of months during a year, but the name based on the first month of a year has become established and is used mainly because it is the month of January that shows the highest returns on world markets, compared to other months. According to the so-called tax-selling, the January effect can be attributed to the sell-off of loss-making securities at the end of a year prompted by investors who seek to create tax-losses to avoid paying the capital gains tax (Chan,1986; Jones & Lee & Apenbrink, 1991). This has its proponents as well as opponents. Jones, Lee & Apenbrink (1991) tested the on the Cowles Industrial Index before and after 1917, when a personal income tax was introduced. The conclusion they arrived at was that whereas the January effect was not significant for the period before 1917, it proved significant for the latter period, thus the January effect was related to income taxation.

606 Jan Luhan, Veronika Novotná, Vladěna Obrov Bhardwaj &Brooks (1992) come to the conclusion that the January effect is more an effect of low-priced securities whose higher returns are, however, absorbed by transaction fees and thus cannot be used for investment strategies. Ritter (1988) focuses on a detailed study of development in particular months and concludes that January does bring higher yields, and they are significantly higher during the first days. 6. Methodology and data The January Effect will be examined in the years 2007 2010 by using official lists of prices. For our research we have chosen a package of stocks traded on the main Prague Stock Exchange market including stocks of the following companies in the period from 2007 to 2010: 1. AAA Auto Group N.V. (AAA) 2. Central European Media Enterprises Ltd. (CETV) 3. Česká zbrojovka, a.s. (ČESKÁ ZBROJOVKA) 4. ČEZ, a.s. (ČEZ) 5. ECM Real Estate Investments A.G. (ECM) 6. Erste Group Bank AG (ERSTE GROUP BANK) 7. Jihomoravská plynárenská, a.s. (JM PLYNÁRENSKÁ) 8. Komerční banka, a.s. (KOMERČNÍ BANKA) 9. New World Resources N.V. (NWR) 10. Orco Property Group S.A. (ORCO) 11. Pegas Nonwovens SA (PEGAS NONWOVENS) 12. Pražská plynárenská, a.s. (PRAŽSKÁ PLYNÁREN.) 13. Setuza, a.s. (SETUZA) 14. Telefónica O2 Czech Republic,a.s. (TELEFÓNICA O2 C.R.) 15. Unipetrol, a.s. (UNIPETROL) 16. VGP NV (VGP) 17. Vienna Insurance Group (VIG) 18. Zentiva N.V. (ZENTIVA). Returns achieved on a given day are examined on the basis of average returns and only in the main Prague Stock Exchange market, which will make it possible for us to avoid any inconsistency with insufficient liquidity as its influence particularly on small and less traded stocks may be essential. For each stock we have identified the average, standard deviation, number of observations, plus t-statistics testing the that the average return of the month is the same as the one of the other months. Returns were monitored both separately for each year of the survey and in aggregate for the entire period of four years. An overview of the results is summarised in the tables below.

January Effect at the Czech Capital Market 607 Month Average return Table 1. Summary of results (Source: Own processing) 2007 SD N t-statistics H 0 (α = 0.05) H 0 (α = 0.01) January 0,2049 1,0571 330 3,4057 rejected rejected February -0,0706 1,5139 314-1,1987 accepted accepted March 0,2754 1,4249 308 3,388 accepted accepted April 0,0896 1,5908 325 0,8211 accepted accepted May 0,0627 1,2944 312 0,5879 accepted accepted June 0,057 0,9043 310 0,7182 accepted accepted July -0,1893 1,5306 315-2,6877 rejected rejected August 0,0113 1,8875 330-0,1249 accepted accepted September -0,0258 1,2155 320-0,7876 accepted accepted October 0,1101 1,371 310 1,2165 accepted accepted November -0,3103 1,7234 325-3,8055 rejected rejected December 0,0636 1,5199 325 0,5227 accepted accepted Month Average return 2008 SD N t-statistics H 0 (α = 0.05) H 0 (α = 0.01) January -0,6873 2,7924 315-2,7395 rejected rejected February -0,0084 1,8898 302 2,8463 rejected rejected March -0,2594 2,1619 308 0,2902 accepted accepted April 0,009 1,5369 325 3,8536 rejected rejected May 0,2353 2,1538 315 4,7418 rejected rejected June -0,5069 2,0881 325-2,0226 rejected accepted July -0,204 2,828 330 0,6179 accepted accepted August 0,032 1,6946 315 3,7035 rejected rejected September -0,8412 5,5603 330-1,9567 accepted accepted October -1,3092 8,7552 315-2,2491 rejected accepted November -0,1818 5,8766 320 0,3665 accepted accepted December 0,2159 3,1246 310 3,123 rejected rejected Month Average return 2009 SD N t-statistics H 0 (α = 0.05) H 0 (α = 0.01) January -0,2521 2,9568 325-3,943 rejected rejected February -1,1442 3,8887 308-7,3105 rejected rejected

608 Jan Luhan, Veronika Novotná, Vladěna Obrov March 0,8325 6,1936 312 1,5302 accepted accepted April 0,9573 3,804 330 3,2123 rejected rejected May 0,271 5,4572 315-0,2474 accepted accepted June -0,0224 2,5278 325-1,3777 accepted accepted July 0,6487 2,8815 330 3,3979 rejected rejected August 0,7213 3,8724 330 2,8999 rejected rejected September 0,1229 2,6343 315-0,2335 accepted accepted October -0,1699 2,4342 310-2,561 rejected accepted November -0,0927 1,8823 315-2,5441 rejected accepted December -0,0167 1,557 330-2,1811 rejected accepted Month Average return 2010 SD N t-statistics H 0 (α = 0.05) H 0 (α = 0.01) January 0,2542 1,8633 320 2,2025 rejected accepted February -0,1566 1,9919 312-1,9396 accepted accepted March 0,3436 1,5067 325 3,9112 rejected rejected April 0,4454 2,3657 330 3,3632 rejected rejected May -0,4502 3,5155 325-2,7857 rejected rejected June -0,403 2,4539 315-3,526 rejected rejected July 0,1809 1,9944 312 1,3237 accepted accepted August -0,1341 2,0702 310-1,652 accepted accepted September 0,301 2,3059 330 2,2089 rejected accepted October -0,0181 1,8125 315-0,6623 accepted accepted November -0,2713 1,8255 305-3,2903 rejected rejected December 0,4354 2,1325 312 3,537 rejected rejected Month Average return 2007 2010 SD N t-statistics H 0 (α = 0.05) H 0 (α = 0.01) January -0,1419 2,3483 1290-1,9376 accepted accepted February -0,3295 2,5162 1236-4,6288 rejected rejected March 0,2932 3,436 1253 3,5853 rejected rejected April 0,3597 2,4934 1310 6,1051 rejected rejected May 0,0277 3,3478 1267 0,621 accepted accepted June -0,2367 2,1118 1275-3,8907 rejected rejected July 0,0836 2,4489 1287 1,7487 accepted accepted

January Effect at the Czech Capital Market 609 August 0,1354 2,4738 1285 2,549 rejected accepted September -0,1245 3,4281 1295-1,1302 accepted accepted October -0,3442 4,76 1250-2,5797 rejected accepted November -0,2214 3,2324 1265-2,3477 rejected accepted December 0,1888 2,1977 1277 3,807 rejected rejected Critical region limit for α = 0,05-1,9634 1,9634 Critical region limit for α = 0,01-2,5829 2,5829 1 Fig. 1. Average monthly stock returns per year (Source: Own processing) 2 Fig. 2. Average stock returns per month (Source: Own processing)

610 Jan Luhan, Veronika Novotná, Vladěna Obrov It is apparent from the overall results in the tables that in 2007, before the financial crisis started, the Czech market showed both the January effect and changes occurring at the beginning of holiday period, i.e. lower returns in June and July. January returns are, on the other hand, the highest of the whole year. 2007 can be seen as a typical year matching the hypotheses. The approaching economic crisis is already reflected in the returns of 2008, so other effects than the January effect can be observed. January returns on shares are negative due to the emerging economic crisis, which became more noticeable towards the end of 2008 with the lowest returns in October. It can also be noted that most returns are negative, so the cannot be accepted if return values are positive. Year 2008 is an atypical year in which the January effect cannot be observed. The beginning of 2009 is still marked by the economic crisis, January returns are still negative and the January effect does not occur. Due to enormous negative returns in the first two months and at the end of 2009 the hypotheses with higher positive average returns are rejected. February saw the lowest returns, while April saw the highest returns. The, however, must be rejected in both cases. Year 2009 is characterized by low returns or big differences in returns during particular months. January 2010 saw a positive return again, one of the highest during the year, while summer months regarded as holiday months saw the biggest drop. In this respect, year 2010 shows signs of working normally but is still partly affected by the economic crisis. Generally, the lowest returns were observed in all months of 2008. As stated above, this can be attributed to the economic crisis, which started on the financial markets. The highest returns of 2009 were observed in April. Conclusion The January effect has not been confirmed in all of the examined periods. If we look at each of the years, it is obvious that only in 2007, we can observe the January effect and further the so-called holiday effect, which may be evoked for example by the summer vacations. In other years 2008-2010 we do not observe the January effect, sometimes just a holiday effect. This is due to the economic crisis, which began to emerge since 2008 as the financial crisis. But it is important to watch over the changes of the average stock returns each year and find a trend within them. It is possible that time will relocate the January effect on our markets, and this will happen when the economic crisis is over and the financial markets once again start to behave in a more standard fashion. References 1. BHARDWAJ, R., BROOKS, L. The January Anomaly: Effects of Low Share Price, Transactions Costs, and Bid-Ask Bias. The Journal of Finance, 1992, Vol. 47, No. 2, s. 553-575.

January Effect at the Czech Capital Market 611 2. CHAN, K. C. Can Tax-Loss Selling Explain the January Seasonal in Stock Returns? The Journal of Finance, 1986, Vol. 41, No. 5, pp. 1115-1128. 3. CHANG, E., PINEGAR, M., RAVICHANDRAN, R. International Evidence on the Robustness of the Day-of-the-Week Effect. The Journal of Financial and Quantitative Analysis, 1993, vol. 28, no. 4, pp. 497-513. 4. FAMA, E. Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal of Finance, 1970, vol. 25, pp. 383-417. 5. FAMA, E. Efficient Capital Markets: II. The Journal of Finance, 1991, vol. 46, no. 5, pp. 1575-1617. 6. FILÁČEK, J., KAPIČKA, M., VOŠVRDA, M. Testování hypotézy efektivního trhu na BCPP. Finance a úvěr, 1998, roč. 48, č. 9, s. 555-566. ISSN 0015-1920. 7. GIBBONS, M., HESS, P. Day of the Week Effects and Asset Returns. The Journal of Business. Chicago : The University of Chicago Press, 1981, vol. 54, no. 4, pp. 579-596. 8. HARRIS, L. A Transaction Data Study of Weekly and Intradaily Patterns in Stock Returns. Journal of Financial Economics, 1986, vol. 16, pp. 99-117. 9. HAUGEN, R. A. Modern Investment Theory. 4th edition. London : Prentice-Hall, 1996. 748 p. ISBN 0-13-261397-2. 10. JONES, S., LEE, W., APENBRINK, R. New Evidence on the January Effect Before Personal Income Taxes. The Journal of Finance, 1991, Vol. 46, No. 5, pp. 1909-1924. 11. LIU, Y. Discussing on Trend to Efficient Market Hypothesis of Securities and Futures Market. In ICIII 09 Proceedings. [Online]. Washington D. C. : IEEE Computer Society, 2009. Vol. 3. p. 149-152. ISBN: 978-0-7695-3876-1. 12. Zhe, L., ZhaoYang, d., Sanderson, P. The Efficient Market Hypothesis and Electricity Market Efficiency Test. 2005. Transmission and Distribution Conference and Exhibition: Asia and Pacific. pp. 1-7. 13. LUHAN, J., NOVOTNÁ, V., OBROVÁ, V. Weekend Effect at Czech Capital Market. 1st international scientific conference Practice and research in private public sector-11. 2011. 1(1). pp. 178-182. ISSN 2029-7378. 14. MALKIEL, B. The Efficient Market Hypothesis and Its Critics. Journal of Economic Perspectives, 2003, vol. 17, no 1, pp. 59-82. 15. RITTER, J. The Buying and Selling Behaviour of Individual Investors at the Turn of the Year. The Journal of Finance, 1988, Vol. 43, No. 3, pp. 701-717. 16. SAMUELSON, P. A. Proof That Properly Discounted Present Values of Assets Vibrate Randomly. The Bell Journal of Economics and Management Science Vol. 4, No. 2 (Autumn, 1973), pp. 369-374. 17. Tsibouris, G., Zeidenberg, M. Back propagation as a test of the efficient markets. In Proceedings of the Twenty-Fifth Hawaii International Conference, 1992, pp. 523 532. Sausio efektas Čekijos kapitalo rinkoje Jan Luhan, Veronika Novotná, Vladěna Obrová Santrauka. Investicijų procese būtina ištirti didelį akcijų skaičių finansinių aktyvų kategorijose. Pagrindinis tikslas yra rasti akcijas, kurios pirkėjo požiūriu esamu metu nėra korektiškai

612 Jan Luhan, Veronika Novotná, Vladěna Obrov įvertintos. Tam naudojami skirtingi požiūriai, tarp jų veiksmingų rinkų teorija. Šiame straipsnyje veiksmingų rinkų teorijos metodai taikomi Prahos vertybinių popierių biržoje 2007 2010 m. ir tikrinamas vad. sausio efekto pasireiškimas. Jan Luhan is an assistant at the Department of Informatics, Faculty of Business and Management, Brno University of Technology. He holds a Master s degree in Corporate Finances and Business and a Bachelor s degree in Tax Advisory. His research interest is in E-business with specialization in web development, database systems, artificial intelligence and strategic management. He is currently completing a PhD study in the field of Company Management and Economics. Jan Luhan Brno (Čekija) technologijos universiteto darbuotojas, korporacinių finansų ir verslo magistras; mokslo tyrimų sritys e.verslas, strateginė vadyba. Veronika Novotná is an Assistant Professor at the Department of Informatics, Faculty of Business and Management, Brno University of Technology. She holds a Master s degree in Mathematics and Economics and a PhD degree in the field of Company Management and Economics. She is an expert in the field of statistic, data mining and artificial Intelligence. Veronika Novotná Brno (Čekija) technologijos universiteto asistentė, gynė disertaciją iš bendrovių vadybos ir ekonomikos; mokslo tyrimų sritys statistika, duomenų vadyba ir dirbtinis intelektas. Vladěna Obrová is a PhD student in the field of Company Management and Economics at the Department of Informatics, Faculty of Business and Management, Brno University of Technology. She holds a Master s degree in Applied Mathematics, the field of study was Mathematics and Economics. She is expert in the field of econometrics, risk management and statistics. Vladěna Obrová Brno (Čekija) technologijos universiteto doktorantė bendrovių vadybos ir ekonomikos srityje, taikomosios matematikos magistrė; mokslo tyrimų sritys ekonometrija, rizikos vadyba, statistika.